3.5.2. Low-Carbon Space Optimization

The FLUS model is used to simulate changes and future patterns of land use, which can aid in spatial optimization and assist decision-making [24]. The model uses the artificial neural network (ANN) to calculate the transition probability of each land type and then integrates the system dynamics model (SD) and the Cellular Automaton model (CA) for further analysis. The self-adaptive inertia and competition mechanism are used in the CA model to deal with complex competition and interactions between different land-use types [25]. In this study, the low-carbon land-use allocation plan was used as the quantitative target for land-use type change in the FLUS model. After that, with the consideration of planning targets in the study area's territorial space, we added the following optimization-oriented driving factors (Table 2): First, the probability of land around the main axis of the Jinhua-Yiwu metropolitan area converting to construction land was improved by taking the Jinhua-Yiwu Expressway as a driving factor in order to promote the intensive use of land resources. Second, the hilly area in the central basin was taken as a constraint in order to create a green ecological zone in the central basin area. Through the above method, the spatial pattern of land use after low-carbon optimization was simulated.

**Table 2.** Optimization-oriented driving factors.


#### **4. Results**
